Assessment Model for the Adjustable Potential of Silicon Carbide Enterprises Based on Dynamic Electricity Pricing and Demand Response

With the development of information technology and the construction of new power systems, it has become increasingly important for industrial users with high load demands and complex adjustable potential to actively participate in demand response. Silicon carbide (SiC) enterprises serve as a representative example of such industries, and analyzing their adjustable potential is of general relevance. Properly accounting for revenue and planning production schedules in advance to participate in demand response is crucial. This paper presents a comprehensive adjustable potential evaluation model based on data processing, prediction, and analysis, considering the production process behavior patterns of SiC enterprises. First, a process flow analysis of the SiC enterprise load is conducted to derive the corresponding load system indicators. Next, the obtained load data undergo variational modal decomposition (VMD), incorporating additional features such as the load of other equipment and weather factors. These features are then input into an optimized Long Short-Term Memory (LSTM) network via the Sparrow Search Algorithm (SSA) for the prediction of different components, thereby enhancing the accuracy of the prediction model. Finally, based on the distinct characteristics of the load and the gradual implementation of time-of-use electricity pricing policies, a comprehensive cost model is developed for optimal regulation and interruption strategies. Using the predicted data, the adjustable potential of the enterprise is analyzed. The proposed method is demonstrated through real-world data from a specific enterprise in a certain region. The comparison with different models highlights the advantages and effectiveness of the proposed approach, ultimately showcasing the results of the adjustable potential analysis.